Production AI Institute · Public record

Wednesday 8 July 2026 — Production AI Record

Continue today's briefing

Tracked now: 24 incidents, 19 disclosures, 5 critical. Most repeated risk: Observability. Latest evidence: 8 July 2026.

Observed records

Today's ledger

Source-backed records with dates, confidence, and links into the evidence trail.

Incident
Anthropic: Elevated errors for some models
7 July 202695%2 sources
Anthropic reported elevated errors on requests to some models, primarily Claude Fable 5. Success rates returned to normal but the issue was not fully resolved, and components were in a partial outage state as of the latest update. Affected services included claude.ai, Claude API, Claude Code, and Claude Cowork.
→
Incident
Supabase: Project status change failures in multiple regions
4 July 202695%2 sources
Compute capacity degradation affecting project restart and resize operations across 16 regions since June 30. Error rates are trending down with improved availability for smaller compute sizes. Upgrading to Postgres 17.6.1.121 or higher provides access to additional machine types.
→
Incident
JetBrains: IDE plugin supply-chain key theft pattern
19 June 202684%2 sources
Malicious IDE plugins exfiltrating API keys highlight a toolchain gap in MSP and enterprise AI security reviews.
→
Incident
Y Combinator: Autonomous agent bankruptcy — DN42 operator cost overrun
18 June 202684%2 sources
Forensic record of an autonomous agent deployment that exhausted operator funds when spend governance and circuit breakers were absent.
→
Incident
Google: Network latency and packet loss in India (Delhi, Chennai, Mumbai)
4 July 202695%2 sources
A fire at a third-party data center facility in Delhi caused emergency power shutdown of networking equipment, isolating a Point of Presence and reducing network capacity. Hybrid Connectivity, VPC, and Media CDN customers experienced intermittent latency spikes and packet loss from June 5 to June 26, 2026. Service was restored after capacity augmentation and rerouting.
→
Incident
OpenAI multi-service outage (2026-05-29)
29 May 202686%3 sources
Production AI Institute incident record tracking a multi-service outage affecting OpenAI API consumers on 2026-05-29. See referenced sources for the official status timeline; PSF impact is mapped to provider fallback and observability domains.
→
Incident
OpenAI Multi-Service Outage (ChatGPT, Login, Access)
29 May 202680%2 sources
On 29 May 2026 OpenAI's public status history recorded four separate incidents the same day affecting ChatGPT conversations, login and account creation, ChatGPT access, and business subscription checkout. Teams routing production workloads through a single OpenAI dependency experienced correlated f…
→
Incident
Anthropic: Elevated errors on Claude Opus 4.8
8 July 2026100%2 sources
Elevated errors on Claude Opus 4.8 affected claude.ai, Claude API, Claude Code, and Claude Cowork. The incident started at 2026-07-08 18:00 UTC and was under investigation, with minor impact.
→
Incident
OpenAI: FedRAMP workspace functionality degradation
4 July 202695%2 sources
Multiple features in FedRAMP workspaces are experiencing issues: Codex, workspace analytics, conversation search, custom GPT search, ChatGPT user invites, and Compliance Logs Platform download endpoint. Core functionality has been restored but known ongoing issues remain under investigation.
→
Incident
Reuters: Binnall Law Claude Console Phantom Citations in Federal Court
1 Jan 202680%2 sources
Federal counsel used Anthropic Claude Console to draft a May 2026 motion that included quotations not found in the cited cases. The firm apologised to the court and committed to citation verification via established legal research platforms.
→
Incident
Reuters: Law Firm AI Hallucinated Fake Case Law in Court Filings
1 Jan 202580%2 sources
Reuters reported that major law firms were warning lawyers that AI can invent fake case law and that filing hallucinated citations can trigger sanctions. The issue moved from isolated embarrassment to a repeatable legal-risk pattern as generative AI use spread through litigation teams.
→
Incident
Zillow iBuying Algorithm Collapse
1 Jan 202180%2 sources
Zillow's algorithmic home-buying programme (Zillow Offers) purchased homes at prices its model predicted would appreciate, then resold them for profit. The model failed to track rapidly changing market conditions in 2021. Zillow found itself holding 9,800 homes worth $2.7 billion, purchased at abov…
→
Snapshot generated 8 July 2026 · observed 12h ago. Corrections · How records are made
Open record explorer →
Public record

This record is maintained by PAI and free to cite. If something is wrong or missing, tell us. Corrections and source suggestions keep the record honest.

Get record updates →Submit a correction
Records are free to cite. citation guidance.